if (!require("pacman")) install.packages("pacman")

# pacman must already be installed; then load contributed
# packages (including pacman) with pacman
pacman::p_load(pacman, rio, tidyverse,DataExplorer,viridis,hrbrthemes,skimr,plotly,DT,scales,ggrepel)
# pacman: for loading/unloading packages
# rio: for importing data

df_canada_pr <- import("http://www.cic.gc.ca/opendata-donneesouvertes/data/IRCC_M_PRadmiss_0002_E.xls",skip = 3)

df_canada_pr <-  df_canada_pr%>% 
as_tibble()

colnames (df_canada_pr) [1]<- "Country of citizenship"

df_canada_pr1 <- df_canada_pr %>% 
slice(-1:-1) %>% 
select("Country of citizenship","2015 Total","2016 Total","2017 Total","2018 Total","2019 Total","2020 Total") 

library(lubridate)
df_canada_clean <- df_canada_pr1 %>% 
  gather(key = "yearly_pr",value,starts_with("2"),-"Country of citizenship",na.rm = TRUE) %>% 
  mutate(value=ifelse(value =="--", 0, value)) %>% 
  mutate(value = as.numeric(value)) %>%
  mutate(yearly_pr = str_remove(yearly_pr," Total"))

df_canada_clean %>% 
  filter(`Country of citizenship`!= "Total") %>% 
  group_by(`Country of citizenship`) %>% 
  summarise(total_pr_granted = sum(value), mean = round(mean(value),0), .groups = 'drop') %>% 
  mutate( percent_weight = round(total_pr_granted / sum(total_pr_granted) *100,2)) %>% 
  arrange(desc(total_pr_granted)) %>% 
datatable(rownames = FALSE, filter="top")

df_canada_clean %>% 
  # filter(`Country of citizenship`!= "Total") %>% 
  group_by(`Country of citizenship`,`yearly_pr`) %>% 
  summarise(total_pr_granted = sum(value), .groups = 'drop') %>% 
  mutate(percent_increase_cases = round((total_pr_granted/lag(total_pr_granted) - 1 )  * 100,2)) %>% 
  arrange(desc(yearly_pr)) %>% 
datatable(rownames = FALSE, filter="top")

df_pr_line <- df_canada_clean %>% 
  filter(`Country of citizenship`!= "Total" & yearly_pr != "2020") %>% 
  group_by(`Country of citizenship`,`yearly_pr`) %>% 
  summarise(total_pr_granted = sum(value), .groups = 'drop') %>% 
  arrange(desc(total_pr_granted)) %>%
  top_n(30) %>% 
ggplot( aes(x=yearly_pr, y=total_pr_granted , group=`Country of citizenship`, color=`Country of citizenship`)) +
    geom_line() +
   scale_color_brewer(palette = "Spectral") +
    ggtitle("Yearly PR Granted") +
    theme_ipsum() +
    ylab("pr granted")

ggplotly(df_pr_line)

library(viridis)
library(hrbrthemes)
df_pr_box <- df_canada_clean %>% 
  filter(`Country of citizenship`!= "Total" & yearly_pr != "2020") %>% 
  group_by(`Country of citizenship`,`yearly_pr`) %>% 
  summarise(total_pr_granted = sum(value), .groups = 'drop') %>% 
  arrange(desc(total_pr_granted)) %>%
filter(total_pr_granted > 8000) %>% 
  mutate(`Country of citizenship` = fct_reorder(`Country of citizenship`,total_pr_granted)) %>% 
ggplot(aes(x = `Country of citizenship`, y=total_pr_granted, fill='Country of citizenship')) + 
    geom_boxplot(alpha=0.3) +
    theme_ipsum() +
    theme(legend.position="none") +
    xlab("")+
   scale_fill_brewer(palette="BuPu")+
  expand_limits(y = 0)

ggplotly(df_pr_box)

library(viridis)
library(hrbrthemes)
df_pr_bar <- df_canada_clean %>% 
  filter(`Country of citizenship`!= "Total" & yearly_pr != "2020") %>% 
  group_by(`Country of citizenship`,`yearly_pr`) %>% 
  summarise(total_pr_granted = sum(value), .groups = 'drop') %>% 
  mutate(country_india = fct_other(`Country of citizenship`, keep = "India")) %>% 
  ggplot(aes(x=country_india, y=total_pr_granted)) +
  geom_point() +
  theme_ipsum()+
  scale_color_brewer() +
  theme(legend.position="none")
    
    

ggplotly(df_pr_bar)

df_pr_faceted <- df_canada_clean %>% 
  filter(`Country of citizenship`!= "Total" & yearly_pr != "2020") %>% 
  group_by(`Country of citizenship`,`yearly_pr`) %>% 
  summarise(total_pr_granted = sum(value), .groups = 'drop') %>% 
  arrange(desc(total_pr_granted)) %>%
 filter(total_pr_granted > 5000) %>% 
  mutate(`Country of citizenship` = fct_reorder(`Country of citizenship`,total_pr_granted)) %>% 
  ggplot( aes(x=yearly_pr, y=total_pr_granted, group=`Country of citizenship`, color=`Country of citizenship`)) +
   geom_area() +
    scale_fill_viridis(discrete = TRUE) +
    theme(legend.position="none") +
    ggtitle("PR Granted by each country is last 5 years") +
    theme_ipsum() +
    theme(
      legend.position="none",
      panel.spacing = unit(0, "lines"),
      strip.text.x = element_text(size = 8),
      plot.title = element_text(size=13)
    ) +
    facet_wrap(~`Country of citizenship`, scale="free_y")

ggplotly(df_pr_faceted)

library(viridis)
library(hrbrthemes)
library(ggridges)
 df_canada_clean %>% 
  filter(`Country of citizenship`!= "Total" & yearly_pr != "2020") %>% 
  group_by(`Country of citizenship`,`yearly_pr`) %>% 
  summarise(total_pr_granted = sum(value), .groups = 'drop') %>% 
  filter(total_pr_granted > 5000) %>% 
ggplot(aes(x = total_pr_granted, y = `Country of citizenship`, fill = `Country of citizenship`)) +
  geom_density_ridges() +
  theme_ridges() + 
  theme(legend.position = "none")

library(hrbrthemes)
library(ggridges)
df_pr_scatter <- df_canada_clean %>% 
  filter(`Country of citizenship`!= "Total" & yearly_pr != "2020") %>% 
  group_by(`Country of citizenship`,`yearly_pr`) %>% 
  summarise(total_pr_granted = sum(value), .groups = 'drop') %>% 
  arrange(desc(total_pr_granted)) %>%
 filter(total_pr_granted > 10000) %>% 
  ggplot(aes(x=yearly_pr, y=total_pr_granted ,color = `Country of citizenship`)) +
  geom_line() +
    geom_point()+
    theme_ipsum()
ggplotly(df_pr_scatter)

library(ggrepel)
library(hrbrthemes)
df_pr_text <- df_canada_clean %>% 
  filter(`Country of citizenship`!= "Total" & yearly_pr != "2020") %>% 
  group_by(`Country of citizenship`,`yearly_pr`) %>% 
  summarise(total_pr_granted = sum(value), .groups = 'drop') %>% 
  arrange(desc(total_pr_granted)) %>%
 filter(total_pr_granted > 10000) %>% 
  ggplot(aes(x=yearly_pr,y=total_pr_granted ,color = `Country of citizenship`)) +
  geom_point()+
  geom_text(aes(label=`Country of citizenship`), 
    nudge_x = 0.25, nudge_y = 0.25, 
    check_overlap = TRUE
  ) +
    theme_ipsum()
ggplotly(df_pr_text)